Schrödinger's cat contradiction

[ronenfe]

Hi, i didn't understand why we can't explain the situation simply as the cat is either dead or alive immediately after the poison was released or not, and not to assume that only when the observer opened the box the state of the cat was determined, what assumption does it contradict?

 

[Runner 1]

See, that's the absurdity that Schrödinger was trying to illustrate! Little did he realize is everyone would take on the position that the cat is simultaneously dead and alive.

Experimentally, you can't say ANYTHING about that particular cat until you open the box and "measure" the result.

You can however say that statistically, if you repeated this experiment many times, you are will get "alive" a certain percentage of the time and "dead" the remaining percentage. You can also describe the standard deviation of these measurements. I believe that is all that quantum mechanics will tell you though.

 

[ronenfe]

But why we can't say the same thing on quantum particles, that things happens immediately and not when we measure them.

 

[Runner 1]

How would we set up an experiment to test that?

 

[ronenfe]

Just use a transparent box and watch.

 

[cbetanco]

Just use a transparent box and watch.

That is equivalent to making a measurement. In oder to see the cat (or particle) in the box, you would need to shine light on it so that the light can scatter off the cat and go through the box into yuou eye. So, the shining of the light (which you need to do in order to see the cat or particle or whatever you want to measure) would be equivalent to taking the measurement, and your wave function would collapse into a definite state. So just by seeing the cat, you have unknowingly taken a measurement.

But all this aside, wavefunction collapse is really just an illusion created by two interacting quantum system (in this case the cat in the box and you, the observer with the measureing apparatus), and the whole closed quantum system (you plus cat plus rest of the universe) evlovles in a predictable way according to Schrödinger's equation. Now, if we could only figure out the infinite dimemsional state vector of the universe at a given time...

 

[ronenfe]

Ok, but still why do we need to claim that before seeing the cat he can be both dead and alive and not dead or alive. Also you don't have to look on the cat, you can hear him breathing and know his state.

 

[Ibix]

What you are describing is called a 'local hidden variable theory'. That means a theory in which the cat is alive or dead, but the information is hidden from us (by the box, in this case). The word 'local' comes into it because you haven't invoked any global information such as many-worlds theory.

Incidentally, the 'hidden' bit also explains why building a transparent box doesn't help. Nobody disputes that the cat is definitely alive or definitely dead when we are looking. It's what happens when we aren't looking that is under debate and, with a transparent box, we are always looking.

Back to the question - in the 1980s, a chap called John Bell developed the Bell Inequality. Basically, this shows that for a particular two measurements that you can make, one will on average be less than the other if any local hidden variable theory were true. Put simply, you can design a simple gambling game using quantum particles instead of dice. Bell's inequality says that in a universe like you describe, there's a System for playing this game; in a quantum universe the house always wins. Experiments have shown that the house always wins (this is the real world, after all), so local hidden variable theories do not accurately describe the way the world works.

If the idea of an alive-and-dead cat bothers you, there are three options.

1) Deal with it.
2) Many-worlds theory, where the universe splits in two when the box closes. The cat is dead in one and alive in the other; the two universes stay 'in touch' until the box opens when they permanently and irrevocably part ways and you can find out which one you're in. This is a global hidden variable theory - the variable is which universe you are in, and a better definition of 'not local' is hard to come by.
3) Pilot Wave theory, aka de Broglie-Bohm theory. This is another global hidden variable theory, where the hidden variable is a universe-wide, impossible to detect, but nevertheless real 'pilot wave' which guides particles around while leaving the cat apparently dead-and-alive to anyone not looking.

That last reads more skeptical than I intended. There are serious physicists who hold each of these positions. I'm strictly an amateur these days, so withhold judgment. That also means that there's a fourth option, which is 'something I am not aware of'.

Hope that helps.

PS: It's the finding out that the cat is alive or dead that is important, not how you do it. Hearing breathing, monitoring oxygen consumption, and opening the box for a look are all the same in this context.

PPS: Lunchtime reading tells me that Bell's work was done in the 1960s, not the 1980s, and that there are still loopholes in the tests of Bell's Inequality that might permit local hidden variable theories to exist under certain circumstances. How improbable these circumstances are I am unable to judge, but consensus seems to be that the preponderance of evidence is for quantum theory rather than local hidden variables.

 

[Runner 1]

What you are describing is called a 'local hidden variable theory'. That means a theory in which the cat is alive or dead, but the information is hidden from us (by the box, in this case). The word 'local' comes into it because you haven't invoked any global information such as many-worlds theory.

Incidentally, the 'hidden' bit also explains why building a transparent box doesn't help. Nobody disputes that the cat is definitely alive or definitely dead when we are looking. It's what happens when we aren't looking that is under debate and, with a transparent box, we are always looking.

Back to the question - in the 1980s, a chap called John Bell developed the Bell Inequality. Basically, this shows that for a particular two measurements that you can make, one will on average be less than the other if any local hidden variable theory were true. Put simply, you can design a simple gambling game using quantum particles instead of dice. Bell's inequality says that in a universe like you describe, there's a System for playing this game; in a quantum universe the house always wins. Experiments have shown that the house always wins (this is the real world, after all), so local hidden variable theories do not accurately describe the way the world works.

If the idea of an alive-and-dead cat bothers you, there are three options.

1) Deal with it.
2) Many-worlds theory, where the universe splits in two when the box closes. The cat is dead in one and alive in the other; the two universes stay 'in touch' until the box opens when they permanently and irrevocably part ways and you can find out which one you're in. This is a global hidden variable theory - the variable is which universe you are in, and a better definition of 'not local' is hard to come by.
3) Pilot Wave theory, aka de Broglie-Bohm theory. This is another global hidden variable theory, where the hidden variable is a universe-wide, impossible to detect, but nevertheless real 'pilot wave' which guides particles around while leaving the cat apparently dead-and-alive to anyone not looking.

That last reads more skeptical than I intended. There are serious physicists who hold each of these positions. I'm strictly an amateur these days, so withhold judgment. That also means that there's a fourth option, which is 'something I am not aware of'.

Hope that helps.

PS: It's the finding out that the cat is alive or dead that is important, not how you do it. Hearing breathing, monitoring oxygen consumption, and opening the box for a look are all the same in this context.

PPS: Lunchtime reading tells me that Bell's work was done in the 1960s, not the 1980s, and that there are still loopholes in the tests of Bell's Inequality that might permit local hidden variable theories to exist under certain circumstances. How improbable these circumstances are I am unable to judge, but consensus seems to be that the preponderance of evidence is for quantum theory rather than local hidden variables.

I've got a question. This describes an ensemble of multiple measurements. Which says nothing about one particular case for one particular cat. As far as I can tell, no experiment can be performed on one instance of a cat to determine its state because that would HAVE to involve a measurement. It's meaningless to talk about just one cat-in-a-box.

So maybe it's a semantics thing, but to me, it is incorrect to say "this particular cat is in a superposition of states until measured". It would make more sense to say "nobody knows what state this particular cat is in, but if we perform this experiment many times with the same set-up, we can statistically describe what will happen".

Thoughts?

 

[phinds]

So maybe it's a semantics thing, but to me, it is incorrect to say "this particular cat is in a superposition of states until measured". It would make more sense to say "nobody knows what state this particular cat is in, but if we perform this experiment many times with the same set-up, we can statistically describe what will happen".

Thoughts?

I think that the thing that both you and the OP are hung up on is that the cat is a macro object and the thought of a macro object being in superposition doesn't seem to make sense (doesn't make sense to me, that's for sure).

BUT ... the heart of the experiment is NOT the cat, it's the quantum decay. THAT is what's in superposition ... the cat being dead or alive is a result of the actual state of the quanta. Schrodinger was using the cat to exaggerate the weirdness of superposition and the fact that we don't KNOW the state of the quanta until we measure it. That this leads to the concept of the cat being both alive and dead was, again, just to exaggerate the weirdness.

I'm in the group that says that the cat is always EITHER alive or dead but not both, and statistically, if you open the box really soon, the cat's more likely to be alive and if you wait a really long time to open the box, the cat is likely to be dead (and this is aside from the fact that nobody' thought to FEED the damn thing!). BUT ... you have to make the measurement to KNOW whether it's dead or alive and until you measure it, you can't know.

 

[San K]

Hi, i didn't understand why we can't explain the situation simply as the cat is either dead or alive immediately after the poison was released or not, and not to assume that only when the observer opened the box the state of the cat was determined, what assumption does it contradict?

it can be explained simply...thus:

the decay process/particle maybe in a quantum superposition but the cat is not.

the cat is separate in the sense that first a determinate state of the particle (i.e wavefunction collapse) has to happen...till then the cat is alive.

this is also the solution to the apparent "paradox" re: Schrödinger's cat

the cat is never linked to the quantum state of the particle (on a simultaneous basis, i.e. without time-space separation). its only after the wavefunction collapses is the poison released or not released

the experiment is a series of steps/processes...the cat is removed at least one or two "steps/processes/events/space-time separation" away from the wave-function of the particle...

the state of the cat was never (and cannot be) "integrated" with the wavefunction to being with...thus the cat's health (dead or alive) is never in superposition...we might not know the state of the cat (simply because we made the apparatus conditions thus...:)...) ...but that does not mean it's in superposition of being both dead and alive

 

[Runner 1]

I'm in the group that says that the cat is always EITHER alive or dead but not both, and statistically, if you open the box really soon, the cat's more likely to be alive and if you wait a really long time to open the box, the cat is likely to be dead (and this is aside from the fact that nobody' thought to FEED the damn thing!). BUT ... you have to make the measurement to KNOW whether it's dead or alive and until you measure it, you can't know.

That's exactly* what I'm saying (minimal statistical interpretation) -- I don't see where we are differing.

*Actually, I support the slightly stronger statement that the cat is "alive, dead, or something else". I mean, (and to further weird-ify Schrodinger's example), let's assume the cat is "really" a frog until it's observed. An experimenter would never know the difference.

The approach I am taking is "here are the statistics -- we can use these to predict things over many measurements. Let's not guess at other things we can't ever find out".

 

[skippy1729]

If the idea of an alive-and-dead cat bothers you, there are three options.

1) Deal with it.
2) Many-worlds theory, where the universe splits in two when the box closes. The cat is dead in one and alive in the other; the two universes stay 'in touch' until the box opens when they permanently and irrevocably part ways and you can find out which one you're in. This is a global hidden variable theory - the variable is which universe you are in, and a better definition of 'not local' is hard to come by.
3) Pilot Wave theory, aka de Broglie-Bohm theory. This is another global hidden variable theory, where the hidden variable is a universe-wide, impossible to detect, but nevertheless real 'pilot wave' which guides particles around while leaving the cat apparently dead-and-alive to anyone not looking.

There is, of course, a fourth option:

The quantum state is only a subjective encapsulation of the knowledge that the observer has of the quantum system.
The quantum system has a physical reality; the quantum state does not. The discontinuous change in the quantum state upon measurement is merely the observer updating HIS quantum state using HIS knowledge of the result of the measurement. This neatly solves the apparent difficulties of Schroedinger's Cat and also Wigner's Friend.

Skippy

 

[ronenfe]

it can be explained simply...thus:

the decay process/particle maybe in a quantum superposition but the cat is not.

the cat is separate in the sense that first a determinate state of the particle (i.e wavefunction collapse) has to happen...till then the cat is alive.

this is also the solution to the apparent "paradox" re: Schrödinger's cat

the cat is never linked to the quantum state of the particle (on a simultaneous basis, i.e. without time-space separation). its only after the wavefunction collapses is the poison released or not released

the experiment is a series of steps/processes...the cat is removed at least one or two "steps/processes/events/space-time separation" away from the wave-function of the particle...

the state of the cat was never (and cannot be) "integrated" with the wavefunction to being with...thus the cat's health (dead or alive) is never in superposition...we might not know the state of the cat (simply because we made the apparatus conditions thus...:)...) ...but that does not mean it's in superposition of being both dead and alive

That's the most logic thing to assume, that's the reason why i asked why to ever think different.
Also about setting an experiment, why do we have to set an experiment to prove the cat is either dead or alive why don't you set an experiment to prove the cat is both dead and alive, i mean i can describe a live cat or a dead cat, can you describe a cat that is both dead and alive?

also if we go back to quantum size, i want to claim that also the particle is either decayed or not and we just don't know what state it is, it has 50 50 chance to be in either state, still it is not in a superposition of both decayed and not at the same time.

about multiple universes , that's even more odd than dead-alive cat, i mean where exactly are all the other universes located?

 

[unusualname]

Schrödinger Cat type experiments have actually been realized experimentally (though obviously not with an actual cat). The main problem with macroscopic objects is preventing decoherence destroying the macroscopic superposition, so these experiments aren't quite at macroscopic scale yet:

http://physicsworld.com/cws/article/news/2815
http://www.nature.com/news/2010/100317/full/news.2010.130.html
http://www.nature.com/news/2011/110405/full/news.2011.210.html

But there are proposals to go further eg with mini-mirrors:

http://www.nature.com/news/2003/031001/full/news030929-3.html

and even a virus:

http://www.nature.com/news/2009/090910/full/news.2009.903.html

 

[Ibix]

I've got a question. This describes an ensemble of multiple measurements. Which says nothing about one particular case for one particular cat. As far as I can tell, no experiment can be performed on one instance of a cat to determine its state because that would HAVE to involve a measurement. It's meaningless to talk about just one cat-in-a-box.

So maybe it's a semantics thing, but to me, it is incorrect to say "this particular cat is in a superposition of states until measured". It would make more sense to say "nobody knows what state this particular cat is in, but if we perform this experiment many times with the same set-up, we can statistically describe what will happen".

Thoughts?

Statistics are the summary of the behaviour of all the individuals being observed. In this context, that means that Bell showed that if individual cats behave as you describe then ensemble behaviour in tests of his work would satisfy his inequality. The observed ensemble behaviour does not obey Bell's Inequality, so the individual behaviour cannot be as you describe.

Put another way, Bell showed that the statements "we don't know which state it's in" and "it actually is not in either state" yield different statistical behaviours in certain cases. Experiment rules out the first statement. This does not mean the second one is true, but it's predictions are very very good so it can't be far off.

I agree that there are a lot of issues with Schrodinger's Cat. Can you really put macroscopic objects into superposed states? If you can, doesn't the cat count as an observer? What about Wigner's friend? But these are topics under investigation - as per the links in the post above - and boil down to: "exactly how strange is this stuff?" and not: "is it strange?"

 

[Ibix]

That's the most logic thing to assume, that's the reason why i asked why to ever think different.
Also about setting an experiment, why do we have to set an experiment to prove the cat is either dead or alive why don't you set an experiment to prove the cat is both dead and alive, i mean i can describe a live cat or a dead cat, can you describe a cat that is both dead and alive?

also if we go back to quantum size, i want to claim that also the particle is either decayed or not and we just don't know what state it is, it has 50 50 chance to be in either state, still it is not in a superposition of both decayed and not at the same time.

about multiple universes , that's even more odd than dead-alive cat, i mean where exactly are all the other universes located?

Bell showed that there were consequences to the statement "the particle is either decayed or not and we just don't know what state it is", and that there were different consequences to the statement that the atom is both decayed and not decayed. Experiment is not consistent with the consequences of the first statement, so it is not true, whether or not you find it more logical in the abstract.

As per the links in unusualname's post, experiments continue to see whether or not you can actually superpose a cat. But quantum-scale objects are apparently superposable, and definitely not one-or-the-other.

 

[Ken G]

Another point I would add to the above is that to me, the question here comes down to "what do you want quantum mechanics to be." If you only want it to be a way to make statistical predictions that work, you can use the minimalist "ensemble interpretation" described by Runner_1-- then you don't use quantum mechanics to say anything about an individual system, it is only meant for statistical predictions of ensembles. But, many ask, "is that all quantum mechanics can be? What happens if we try to use it to be our best language for talking about the state of a single system?" In other words, do we want quantum mechanics to be a way of talking about what is really happening to a single system, or don't we?

This is where all the other interpretations come in. We can all agree that we don't test these interpretations, we only test the quantum mechanical predictions themselves-- the interpretations are simply what we make of those predictions, and the only test they need to pass is self consistency. Many debate the self-consistency of these interpretations, but by and large they all seem to be.

A key point to bear in mind, if one is to leave behind the ensemble interpretation and penetrate further into the possibilities, is that quantum mechanics obeys a principle that it still works on large systems of particles and on macroscopic scales, you just wouldn't use it for that-- it's hitting a nail with a sledgehammer. But it does work, so this gives it promise as being a "fundamental" description (whereas the classical mechanics that also works on those systems does not work in general on individual particles on small enough scales). The ensemble interpretation essentially gives up on trying to take advantage of this feature-- we can take a single particle, ramp up its "action" until it is huge (like an elementary particle in a bubble chamber), and treat it classically, without needing to imagine it is just one possibility in an ensemble-- we can know about the particle. Many want quantum mechanics to work like that also, when we consider actions of order h and not just huge actions.

There is a price to pay for this desire-- we encounter fundamental limits about what we can say about systems like that. So do we give up on the whole endeavor, because of those limits, or do we press on anyway? If we press on, we must face problematical issues around the cat paradox and the transition between our goal and the state of a macroscopic system. Something has to give-- you have to put a "cut" somewhere, a cut between what you can know and what you might wish you could know. The ensemble approach, as we've seen, places the "cut" between an ensemble and a single system-- we can know about the former but not the latter. The Copenhagen approach puts in the "Heisenberg cut", which is a cut between the macroscopic worlds where we, as macro observers, do physics, and the microscopic world which is seen only dimly (and which indeed Bohr said doesn't even exist in the same way as the macro world). That is a different way to put in the cut, because it is based on the amount of "action", even for a single particle, rather than on the number of particles in the ensemble.

Then we come to more bizarre ways to put in the "cut", like the many-worlds approach, which places the cut encircling the coherent island that we actually experience and acknowledge as a subset of a full reality that we do not perceive and most wouldn't acknowledge at all-- but this allows us to talk about the state of the cat as always being in a superposition of alive and dead even after we observe it and erroneously think is one or the other. And there is the deBroglie-Bohm interpretation, which is in a sense the opposite approach-- it also places the "cut" between what is real and what we can actually know is real, but instead of saying that what is real is a superset of what we can know, it says that what is real is a subset of what we can know-- there is an actual state of the system that is one out of a range of possibilities that we group together into the superposition state, even for a single particle on atomic scales. Hence many-worlds is in some sense the least classical, and deBroglie-Bohm the most classical, with Copenhagen in the middle and the ensemble approach saying only the minimum needed to allow quantum mechanics to be used at all.

Basically, they all seem to me to work at some level, despite the philosophical objections they may raise. The most important thing is that you realize you can't have everything, some fundamental limitation must be embraced. So if the cat dies, by picking an interpretation, at least it gets to choose its poison!

 

[San K]

That's the most logic thing to assume, that's the reason why i asked why to ever think different.
Also about setting an experiment, why do we have to set an experiment to prove the cat is either dead or alive why don't you set an experiment to prove the cat is both dead and alive, i mean i can describe a live cat or a dead cat, can you describe a cat that is both dead and alive?

also if we go back to quantum size, i want to claim that also the particle is either decayed or not and we just don't know what state it is, it has 50 50 chance to be in either state, still it is not in a superposition of both decayed and not at the same time.

about multiple universes , that's even more odd than dead-alive cat, i mean where exactly are all the other universes located?

one way to look at it sh/c/w ould be...great men think alike...;)